Quality control of a product or process has become a large part of the economics of industry. Two major concerns of quality control are accuracy and the time it takes to detect the slightest mistake in a process. Human resources are heavily relied upon for inspection of various aspects of the production process. A typical inspection procedure is to analyze a sampling of products. Various devices are employed to measure differences in weight, temperature and other dimensions. Such devices and procedures, although accurate for the sample, are usually time consuming and are incapable of giving a dynamic quantitative analysis. Subtle and intermittent defects are often undetected.
Radiation detectors have been used as non-contact alternatives to many temperature sensors and detect abnormalities by measuring thermal changes. In particular, infrared sensing devices are being used to detect temperature differences between a subject and a reference, and to measure heat gain or loss from machinery, plumbing, electrical lines and the like. Typically, radiation sensors respond to changes in thermal radiation in the order of less than 1/10 second. Such sensors are not only fast, but are accurate and economic as operations of interest do not have to be shut down during detection.
However, such radiation detectors have not been used to detect a dynamic pattern of temperatures or temperature changes as is found in assembly lines and conveyer belt type processes. Further, some radiation detectors operate from an internal reference which is built into the sensor itself. Accurate interpretation of the detector's signal requires careful calibration. Such detectors are limited by their calibration, and thus cannot detect subtle changes in radiation outside of the limits of the calibration.
Also, radiation detectors often become dirty and overheated in certain work environments. Elaborate colling, purging and cleaning systems have been used, but are expensive, clumsy and require maintenance to retain proper calibration.